Catalytic activity in individual cracking catalyst

0 downloads 0 Views 2MB Size Report
Sep 18, 2011 - Fluid catalytic cracking (FCC) is the major conversion process used in oil refineries ... the performance of FCC catalyst particles, which are complex ..... Cheng, W-C. et al. in Handbook of Heterogeneous Catalysis (eds Ertl, G.,.
ARTICLES PUBLISHED ONLINE: 18 SEPTEMBER 2011 | DOI: 10.1038/NCHEM.1148

Catalytic activity in individual cracking catalyst particles imaged throughout different life stages by selective staining Inge L. C. Buurmans1†, Javier Ruiz-Martı´nez1†, William V. Knowles2, David van der Beek3, Jaap A. Bergwerff3, Eelco T. C. Vogt3 and Bert M. Weckhuysen1 * Fluid catalytic cracking (FCC) is the major conversion process used in oil refineries to produce valuable hydrocarbons from crude oil fractions. Because the demand for oil-based products is ever increasing, research has been ongoing to improve the performance of FCC catalyst particles, which are complex mixtures of zeolite and binder materials. Unfortunately, there is limited insight into the distribution and activity of individual zeolitic domains at different life stages. Here we introduce a staining method to visualize the structure of zeolite particulates and other FCC components. Brønsted acidity maps have been constructed at the single particle level from fluorescence microscopy images. By applying a statistical methodology to a series of catalysts deactivated via industrial protocols, a correlation is established between Brønsted acidity and cracking activity. The generally applicable method has clear potential for catalyst diagnostics, as it determines intra- and interparticle Brønsted acidity distributions for industrial FCC materials.

T

he ever increasing worldwide cracking capacity emphasizes the importance of the FCC process for the production of refined oil products. In 1966 the catalytic cracking capacity was 0.5 million tons per day1. In 2004, this had increased to over 2 million tons per day2. Zeolites have been used in FCC catalyst particles since the 1960s because of their suitable acidic properties3–7. Today, the annual global synthetic zeolite industry is worth about US$1.9 billion8. Because cracking is initiated on the acid sites of FCC particles, the acidity of the FCC components influences the overall cracking activity of the catalyst. Industrially used FCC catalyst particles are complex micrometresized spheres (average diameter, 70 mm) that contain zeolitic material (mostly zeolite Y) as well as matrix components such as clay, alumina and silica. During the cracking process, the activity of the catalyst particles decreases due to coke formation and several irreversible deactivation mechanisms, such as hydrothermal dealumination and metal poisoning. So far, these changes in activity have predominantly been studied on a macroscopic scale by investigating activity and selectivity during catalytic cracking tests9. However, such surveys fall short on giving information about the activity and acidity within as well as between individual catalyst particles. Furthermore, knowledge about the exact structure of FCC catalyst particles is scarce. Studies on the structure of both fresh and poisoned FCC catalyst particles have been performed using microscopic and spectroscopic approaches10–16. Such studies have been conducted mostly using scanning electron microscopy– energy dispersive X-ray spectroscopy, electron probe microanalysis, X-ray photoelectron spectroscopy, atomic force microscopy and secondary ion mass spectrometry. However, analytical tools to study the active sites of the catalyst particles in microscopic detail have so far not been available. The use of fluorescence microscopy in combination with probe molecules is a powerful tool in life sciences research17–22 and has

recently been introduced in the field of heterogeneous catalysis23–30. The ability to stain with fluorescent dyes that selectively bind to individual targets in cells and tissues permits the visualization of cellular components and distinct biological events31. Here, we report the first application of such a staining approach to industrially important FCC catalyst particles to selectively visualize zeolite domains. Based on this methodology, it is possible to correlate fluorescence intensity with Brønsted acidity and catalytic activity at the single particle level for FCC catalyst materials throughout different life stages. As both the active zeolite domains and the catalyst particles themselves are on the scale of micrometres, the spatially resolved acidity of the material can be unravelled using confocal fluorescence microscopy. We strongly believe that our methodology can be applied in a diagnostic manner to pinpoint the acidity of the zeolitic phase in a wide variety of catalyst formulations.

Results and discussion Confocal fluorescence microscopy. With biological staining approaches in mind, a systematic confocal fluorescence microscopy study was designed to differentiate between discrete materials in the FCC catalyst particles. The use of staining approaches has also been explored previously in materials science23,32,33. Our methodology involves the use of two different dyes, as illustrated in Fig. 1a. In the first step, the reactive probe molecule thiophene, which can be converted into fluorescent products over Brønsted acid sites34, was selected to specifically stain the zeolitic particulates. Details of the reaction pathways of the Brønsted acid-catalysed thiophene oligomerization can be found in Supplementary Scheme S135. Visualization of the carbocationic thiophene products after reaction at 373 K was accomplished by illumination with either a 488 nm or 561 nm laser, which cause fluorescence of the species that absorb light at

1 Inorganic Chemistry and Catalysis Group, Debye Institute for NanoMaterials Science, Faculty of Science, Utrecht University, 3584 CG Utrecht, The Netherlands, 2 Research Centre for Catalysts, Albemarle Corporation, Pasadena, Texas 77057-1104, USA, 3 Research Centre for Catalysts, Albemarle Catalysts Company BV, 1030 BE Amsterdam, The Netherlands; † These authors contributed equally to this work. * e-mail: [email protected]

862

NATURE CHEMISTRY | VOL 3 | NOVEMBER 2011 | www.nature.com/naturechemistry

© 2011 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY

ARTICLES

DOI: 10.1038/NCHEM.1148

a

c

b NH N O

S

N

10 μm N

S S

O

2 μm

e

d N

+ HN

10 μm

2 μm

Figure 1 | Schematic of the research approach. a, Confocal fluorescence microscopy is used to visualize distinct components of FCC catalyst particles after staining with two different probe molecules. b,d, Confocal fluorescence microscopy image of a FCC catalyst particle with (b) and without (d) zeolite Y on reaction with thiophene (green) at 373 K (lex ¼ 488 nm, detection 500–550 nm) and subsequent staining with Nile Blue A (red) at 298 K (lex ¼ 638 nm, detection 662–737 nm), intensities all boosted with the same factor. c,e, Magnified views of the highlighted areas in b and d, respectively.

those wavelengths. The reaction temperature was set to accurately initiate the thiophene oligomerization reaction, and was sufficient to ensure the formation of fluorescent probe molecules on Brønsted acid sites. High reaction temperatures were avoided to prevent side reactions, such as coke formation of the initial thiophene oligomerization products. Such side reactions would interfere with the Brønsted acid mapping for which our method was tailored. Second, by using a non-reactive dye, Nile Blue A, which is too large to enter the zeolite micropores, the FCC matrix could be stained. Nile Blue A shows a high fluorescence when excited with a 638 nm laser. The combined application of thiophene and Nile Blue A revealed intraparticle heterogeneities, that is, differences in the position and size of zeolitic domains within the matrix of a single FCC particle. This is illustrated in Fig. 1b–e, which compares two distinct FCC particles with (Fig. 1b,c) and without (Fig. 1d,e) zeolite Y. The particle containing zeolite Y has highly fluorescent green domains with an average size of 2–5 mm, heterogeneously distributed over the redcoloured FCC matrix. The absence of green fluorescence in Fig. 1d,e indicates that the particle without zeolite Y has insufficient acid strength to promote carbocationic thiophene product formation. Our assumption that the thiophene reaction only occurs in the zeolite Y domains was confirmed by a detailed complementary UV-vis spectroscopy study. Time-resolved UV-vis absorption spectra of the separate FCC components, before compounding into a FCC catalyst particle, were taken during the thiophene reaction at 373 K (Supplementary Fig. S1). In the UV-vis spectra of zeolite H-Y, clear absorption bands attributed to the formation of thiophene carbocations are observed. Within the matrix, only the clay was slightly active towards the formation of light-absorbing species. Two batches of FCC catalysts, containing the same zeolite Y but manufactured in different catalyst production plants (denoted as samples FCC 1 and FCC 2 hereafter), were also studied using UV-vis micro-spectroscopy (Supplementary Fig. S2). In addition, they were compared with a FCC catalyst sample in which the zeolitic part was replaced by clay. The results clearly show that the FCC catalyst particles without zeolite are inert to the formation of Brønsted acid-catalysed thiophene products, and no absorption bands are

observed. In contrast, absorption bands arise in the spectra of zeolite-containing FCC particles, corresponding to products formed by the zeolitic domains. Owing to the successful visualization of the zeolite domains within the FCC particles, a three-dimensional (3D) reconstruction of the fluorescence intensity was obtained. Because fluorescent products are formed only at Brønsted acid sites, such an image reveals a 3D map of the Brønsted acidity within the catalyst body. It is known that deactivation in the cracking process is strongly linked to the loss of Brønsted acidity3–7. This opens up the possibility of visualizing Brønsted acidity changes after different deactivation procedures using the thiophene probe reaction. For this purpose, the two catalyst batches under study, FCC 1 and FCC 2, were used as prepared (fresh) and after three industrially relevant deactivation methods: steaming (ST)36, two-step cyclic deactivation (CD)37,38 and Mitchell impregnation–steam deactivation (MI)39. These laboratory techniques attempt to mimic catalyst behaviour in a real cracking unit in terms of deactivation by coke formation, metal deposition and/or hydrothermal ageing. More details regarding the deactivation methods can be found in the Supplementary Information. Fluorescence was detected in the 570–620 nm wavelength range after excitation with a 561 nm laser. Figure 2 shows that, when fluorescence intensities of the different FCC 1 catalyst samples were examined following exposure to thiophene, less fluorescence could be detected for deactivated catalyst samples (trend: fresh . ST . CD . MI). Because sufficiently strong Brønsted acid sites are needed for the formation of fluorescent carbocations, this means that the amount of strong acid sites has decreased following deactivation. After each deactivation procedure, it may be that some of the zeolitic micropores gets blocked. If such a phenomenon occurs, then the formation of fluorescent thiophene oligomerization products can be used to monitor the remaining accessible Brønsted acidity of the FCC particles. The same sequence in fluorescence intensity was measured for the FCC 2 catalyst batch (Supplementary Fig. S3). It is important to mention that deactivation can also influence the accessibility of the catalyst matrix to probe molecules. To rule out this as the main cause of changes in fluorescence intensity, all samples were investigated with confocal fluorescence microscopy after the addition of Nile Blue A. The distribution of Nile Blue A

NATURE CHEMISTRY | VOL 3 | NOVEMBER 2011 | www.nature.com/naturechemistry

© 2011 Macmillan Publishers Limited. All rights reserved.

863

ARTICLES

NATURE CHEMISTRY c

b

10 μm

0.5 0.4 0.3 0.2 0.1

10 μm

10 μm

10 μm

2 μm

2 μm

2 μm

0.6

Relative frequency

0.6

Relative frequency

Relative frequency

2 μm

d

0.5 0.4 0.3 0.2 0.1 0.0

0.0 0

500 1,000 1,500 2,000 Intensity (a.u.)

0

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

Relative frequency

a

DOI: 10.1038/NCHEM.1148

0

500 1,000 1,500 2,000 Intensity (a.u.)

500 1,000 1,500 2,000 Intensity (a.u.)

0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0

500 1,000 1,500 2,000 Intensity (a.u.)

Figure 2 | Confocal fluorescence microscopy methodology for the FCC 1 catalyst batch. a–d, Top row: confocal fluorescence microscopy images of individual FCC particles after reaction with thiophene at 373 K (lex ¼ 561 nm, detection 570–620 nm, false-colour images, intensities all boosted with the same factor) in different stages of deactivation, fresh (a), steamed (b), two-step cyclic deactivated (c) and Mitchell impregnated–steam deactivated (d). Middle row: magnification of the highlighted areas in the top row discloses zeolitic domains in the micrometre size range with decreasing intensity upon deactivation. Bottom row: fluorescence intensity histograms as determined for the zeolite domains in the four different samples.

Statistical Brønsted acidity evaluation. The fluorescent signal of the different samples following reaction with thiophene was studied in detail by performing a statistical analysis of six to eight catalyst particles per sample. It was found that in all studied catalyst particles, both fresh and deactivated, the fluorescence intensities of the zeolite domains were independent of their location within the catalyst body. This observation suggests nonpreferential zeolite domain deactivation within a FCC particle. A more quantitative insight into the fluorescence intensities was acquired by calculating the average fluorescence intensity of all zeolite spots within the population. To resolve intraparticle heterogeneity, a fluorescence intensity threshold was set to remove the FCC matrix fluorescence and the residual fluorescence of zeolite particles in close proximity. Domains that are too close together cannot be resolved with fluorescence microscopy. This phenomenon is described well in the literature for fluorescence imaging of nanoscopic events28,29,40,41. Statistical analysis and threshold setting protocols are explained in more detail in the Supplementary Information (Supplementary Fig. S5). Because, in general, over 90% of zeolitic domains are expected to be larger than 1 mm2 (Albemarle Catalysts, unpublished data), all bright spots below this value were rejected. An average intensity of at least 150 zeolite spots in different catalyst particles was determined for every sample. 864

Fluorescence intensity histograms, showing a population distribution with a clear maximum for all samples (Fig. 2, bottom row), were used to determine the average fluorescence intensities for the different samples. The results of this analysis are given in Fig. 3. To validate our method, a parallel statistical analysis was performed. In this analysis the mean fluorescence intensity per individual FCC particle was used to determine the average fluorescence intensity of the sample. The results of this approach are summarized in Supplementary Fig. S6, and show comparable fluorescence intensity values, which follow the same trends as those described in Fig. 3. The only difference is that the standard errors are significantly larger Error = standard error: s/√n

1,200 Intensity (a.u.)

(Supplementary Fig. S4) was comparable for all samples under investigation. This finding confirms that the catalyst particles are still accessible to probe molecules in the used size range after deactivation.

800

400

0

Fresh

ST

CD

Ecat

MI

Figure 3 | Statistical analysis of the fluorescence microscopy data. Fluorescence mean intensities and their corresponding standard errors, as determined from the statistical analysis of at least 150 zeolitic domains in the confocal fluorescence measurements of FCC catalyst particles: FCC 1 (dark grey), FCC 2 (light grey) and Ecat (black). NATURE CHEMISTRY | VOL 3 | NOVEMBER 2011 | www.nature.com/naturechemistry

© 2011 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY

b 80 70 60 50 40 30

c Fluorescence intensity (a.u.)

Gasoline selectivity (wt/wt)

a 430°F+ conversion (wt%)

ARTICLES

DOI: 10.1038/NCHEM.1148

0.7 0.6 0.5 0.4

1,200

800

400

20 2

3

4 5 Catalyst/oil (wt/wt)

6

7

30

40

50 60 70 Conversion (wt%)

80

0.02 0.04 0.06 Micropore volume (ml g–1)

e

d

f

Signal (a.u.)

Brønsted

Lewis

1,550

1,500 1,450 Wavenumber (cm–1)

1,400

373

473

573 673 Temperature (K)

773

873

10

20

30 40 50 2θ (deg)

60

70

Figure 4 | Validation of the confocal fluorescence methodology. Bulk characterization techniques for the FCC 1 catalyst particles in their fresh (grey), steamed (blue), two-step cyclic deactivated (green) and Mitchell impregnated–steam deactivated (red) state. Black represents the performance of the equilibrium catalyst. a, Activity data. b, Gasoline selectivity. c, Average fluorescence intensity values obtained from thiophene oligomerization on the catalyst particles as a function of the micropore volume determined by nitrogen physisorption. d, Transmission IR-spectra after pyridine adsorption. Bands in d are due to adsorption of pyridine on both Brønsted and Lewis acid sites within the catalyst samples. e, Temperature programmed desorption plots of ammonia. f, X-ray diffraction powder patterns.

due to the smaller number of data points (n ¼ 6 particles versus n ¼ 150 domains). By comparing the fluorescence intensity values of the FCC 1 and FCC 2 catalyst batches, the reliability of the analysis method was tested (Fig. 3). The fluorescence microscopy images for the fresh and deactivated FCC 2 particles (Supplementary Fig. S3) reveal the same trend in average fluorescence intensity decrease, as shown in Fig. 3. The average fluorescence intensity values across the two catalyst types display a maximum difference of 7% (Supplementary Table S1). These observations indicate that our characterization method is very valuable for comparing and evaluating different types of FCC catalyst particles that have experienced distinct deactivation protocols. Link between Brønsted acidity and cracking activity. The question now arises whether mapping of the Brønsted acidity within FCC catalyst particles can be linked to catalytic cracking reactivity. Indeed, a correlation was found between Brønsted acidity and catalytic activity when fluorescence intensities deduced from the confocal fluorescence microscopy measurements were compared with data obtained in laboratory fluid simulation tests (FST) on the different deactivated catalysts for the conversion of vacuum gas oil (VGO). This structure–performance relationship is presented in Fig. 4 for the FCC 1 catalyst materials. Further details about the activity tests as well as results for the FCC 2 catalyst materials are reported in the Supplementary Information (Supplementary Fig. S7). Owing to its high activity, the fresh catalyst could not be tested in such an approach. The deactivated FCC 1 catalyst materials spanned a range of activities. The ST sample, treated under harsh hydrothermal conditions, was the most active. The CD catalyst, subjected to mild hydrothermal deactivation in tandem with metals and coke deposition during cyclic VGO cracking and regeneration, was slightly less active. The MI sample, exposed to moderate hydrothermal conditions after metals

impregnation, showed the lowest conversion. The metals used for CD and MI protocols, nickel and vanadium, promote catalyst deactivation primarily through coke formation and Brønsted acid site destruction. Nominal metal concentrations were kept uniform between CD and MI samples. Therefore, dissimilar conversions between the samples, as reflected in a significantly higher coke formation for the MI sample (Supplementary Fig. S8), hint at fundamental differences in the chemical nature and impact of these metals on the overall catalyst. Both catalytic activity and selectivity towards gasoline, the desired product, is diminished as a consequence of catalyst deactivation (Fig. 4b). When the cracking activity data were compared to the Brønsted acidity changes monitored with our statistical confocal fluorescence analysis, the same trend was observed, suggesting decreasing cracking activity and Brønsted acidity with increasing deactivation severity. As the deactivation techniques we used caused partial loss of the number of Brønsted acid sites and changes in the overall structure of the zeolite, one would expect that the effective number of active sites available for cracking reactions would also decrease. This decrease can be investigated in fine detail using our statistical confocal fluorescence microscopy approach, which greatly helps to explain the observed activity differences in the cracking tests. To corroborate our approach, additional bulk characterization of the FCC catalyst particles was performed. The results, as summarized in Fig. 4, clearly confirm our confocal fluorescence microscopy data and the observed acidity trends. Nitrogen physisorption shows a significant loss of micropore volume across samples, which can be explained by a partial destruction of the zeolite within the catalyst (Fig. 4c). From the literature it is known that cracking activity decreases upon loss of zeolite structure42,43. Pyridine adsorption followed by IR spectroscopy is a well-established technique to reveal acidity in solid materials44. Figure 4d shows that the IR spectroscopy measurements performed on the FCC 1 catalyst particles reveal a gradual decrease in the Brønsted acidity upon deactivation.

NATURE CHEMISTRY | VOL 3 | NOVEMBER 2011 | www.nature.com/naturechemistry

© 2011 Macmillan Publishers Limited. All rights reserved.

865

ARTICLES Deactivation/ageing

NATURE CHEMISTRY

250

650 1,050 1,450 Average fluorescence intensity distribution (a.u.)

1,850

Figure 5 | Interparticle Brønsted acidity mapping. Average fluorescence intensity ranges that cover the intensities of all seven individual FCC 1 catalyst particles in their fresh (grey), steamed (blue), two-step cyclic deactivated (green) and Mitchell impregnated-steam deactivated (red) states. Black represents the average fluorescence intensities as determined for the seven Ecat particles. Colour scheme as in Fig. 4.

Temperature programmed desorption (TPD) of ammonia, depicted in Fig. 4e, reveals an analogous decrease in the overall acidity (total area below the curves), with a loss of strong acidity (peak at 650 K) for the deactivated samples. X-ray powder diffraction (XRD) patterns (Fig. 4f ) of the FCC 1 catalyst particles indicate a loss in crystallinity of the deactivated samples, clearly seen from the increased background contribution. A Le Bail extraction of the measured XRD patterns (see Supplementary Information) reveals a decrease in unit cell size, which is indicative of a loss of framework aluminium atoms from the zeolite, and, as a consequence, a reduction in the quantity of Brønsted acid sites. For the FCC 2 catalyst batch, all results show the same trends, as summarized in Supplementary Fig. S7. Predicting catalyst deactivation in a real FCC unit. We have shown that our confocal fluorescence microscopy approach is very useful in assessing Brønsted acidity for different deactivation stages, and forms a bridge with catalytic activity. The samples studied so far have been deactivated and tested under controlled laboratory conditions to mimic their behaviour in a real FCC unit. As a showcase, the same approach was applied to a so-called equilibrium catalyst, Ecat, taken from a commercial FCC unit. Owing to constant deactivation in the refiner’s FCC unit, degrading catalyst activity is compensated by the continual addition of fresh catalyst to maintain stable activity levels. The resulting Ecat is a mixture of particles with different residence time histories, which therefore display diverse activity levels. Bulk characterization techniques, shown in Fig. 4, provide average information about the Ecat behaviour and indicate that the Ecat cracking activity lies between the activity of the CD and MI samples. This is in good agreement with the acidity trends obtained with pyridine IR spectroscopy and ammonia TPD. The XRD pattern of the Ecat indicates that a significant amount of crystallinity is preserved during the FCC process. A decrease in unit cell size compared with the CD sample illustrates a lower amount of framework aluminium. The average fluorescence intensity obtained with our method allows the Brønsted acidity to be predicted and thus the cracking activity of the Ecat. Figures 3 and 4a confirm that the approach indeed accommodates predictive potential: both the Ecat average fluorescence intensity and cracking activity lie between those of the CD and MI samples. These observations are in line with the Brønsted acidity trends obtained with bulk characterization techniques. As a large heterogeneity in the catalytic activity of individual particles within the Ecat batch is expected, the bulk characterization 866

DOI: 10.1038/NCHEM.1148

techniques are lacking an important piece of information. Fortunately, interparticle Brønsted acidity information within the Ecat batch can be readily obtained with our confocal fluorescence microscopy approach by evaluating the average fluorescence intensity per individual FCC particle. This is depicted in Fig. 5. The range of fluorescence intensities observed for the Ecat sample is wider than for CD and MI combined, reflecting a larger interparticle heterogeneity in terms of age, Brønsted acidity and catalytic activity within an industrial Ecat. These findings clearly show the added value of investigating catalyst batches at the individual particle level. Future research might include the use of single-molecule detection and sub-diffraction spatial resolution methods, which have been applied previously in the study of catalytic materials28,29. This would provide a deeper insight into the activity changes in terms of turnover frequencies within specific zeolite domains of the catalyst particles as a function of catalyst deactivation.

Conclusions It has been demonstrated that confocal fluorescence microscopy allows the visualization of specific catalyst components within individual FCC particles by selective staining, particularly the active zeolitic phase. Thiophene oligomerization, a Brønsted acid catalysed reaction, has proven to be the bridge necessary to translate confocal fluorescence microscopy images into Brønsted acidity maps. The novel characterization method shows sensitivity towards different industrially relevant deactivation protocols and can serve as a powerful tool in the effort to develop enhanced laboratory deactivation procedures that more closely simulate commercial practice. Proper statistical analysis of the confocal fluorescence microscopy data has led to a direct link between fluorescence intensity/Brønsted acidity at the single particle level and catalytic cracking activity of laboratory deactivated bulk catalysts. Furthermore, our method can determine the age distribution of a catalyst in a real FCC unit in terms of interparticle acidity. We expect that the described methodology has potential to deliver new structural and catalytic insights in other industrially relevant catalyst systems.

Methods Thiophene (Merck, .99%), Nile Blue chloride (Acros Organics, pure) and pyridine (Acros Organics, 99þ%) were used as received. Zeolite H-Y (BET surface area ¼ 622 m2 g21), silica (BET surface area ¼ 147 m2 g21), high-crystalline alumina (BET surface area ¼ 34 m2 g21), low-crystalline alumina (BET surface area ¼ 374 m2 g21) and clay (BET surface area ¼ 20 m2 g21) were provided by Albemarle Corporation. Silica and both types of alumina were dried at 373 K for 30 min. All materials were heat-treated at 723 K for 2 h before use to remove any adsorbed species. Two batches of FCC catalyst materials, labelled in the main text as FCC 1 and FCC 2, were also provided by Albemarle Corporation and used as received. Only the CD-deactivated FCC and the Ecat catalyst particles were calcined before use at 973 K for 2 h. UV-vis micro-spectroscopy studies were conducted using an Olympus BX41M upright research microscope provided with a ×10, 0.3 NA objective lens. Illumination of the sample was performed using a 75 W tungsten lamp. The microscopy setup was equipped with a 50/50 double-viewport tube, which accommodated a charge-coupled device (CCD) video camera (ColorView IIIu, Soft Imaging System GmbH) and an optical fibre mount. The microscope was connected to a CCD UV-vis spectrometer (AvaSpec-2048TEC, Avantes) by a 200-mm-core fibre. UV-vis spectroscopy measurements were performed using an in situ cell (Linkam Scientific Instruments FTIR 600) equipped with a temperature controller (Linkam Scientific Instruments TMS 94). The individual FCC components or FCC particles were placed on the heating element of the in situ cell and heated to 373 K for 5 min, after which 15 ml of thiophene was added and UV-vis absorption spectra were taken every 3 s using an acquisition time of 10 ms and averaging the signal 50 times. Confocal fluorescence experiments were performed using a Nikon Eclipse 90i upright microscope with a ×50, 0.55 NA dry objective lens. Confocal fluorescence images were collected by a Nikon A1-SHR A1 R scan head connected to three Melles Griot laser light sources with emission wavelengths of 488 nm (ion laser, 150 mW), 561 nm (yellow diode-pumped solid-state laser, ,50 mW) and 638 nm (diode laser, 150 mW). The emission was detected by a A1-DU4 4 detector unit. The used detection ranges were 500–550 nm, 570–620 nm and 662–737 nm for the 488 nm, 561 nm and 638 nm lasers, respectively. Samples for confocal fluorescence NATURE CHEMISTRY | VOL 3 | NOVEMBER 2011 | www.nature.com/naturechemistry

© 2011 Macmillan Publishers Limited. All rights reserved.

NATURE CHEMISTRY

ARTICLES

DOI: 10.1038/NCHEM.1148

microscopy were prepared by placing FCC catalyst particles on the heating element of the Linkam in situ cell and heating them at 373 K for 5 min. Subsequently, 15 ml of thiophene was added and the heating was stopped after 10 s. Nile Blue A, dissolved in ethanol, was added to the catalyst particles at room temperature.

27.

28.

Received 6 May 2011; accepted 15 August 2011; published online 18 September 2011

29.

References 1. Brouwer, L. E. L. The Petroleum Handbook (Balding & Mansell Limited, 1966). 2. Cheng, W-C. et al. in Handbook of Heterogeneous Catalysis (eds Ertl, G., Kno¨zinger, H., Schu¨th, F. & Weitkamp, J.) 2741–2778 (Wiley-VCH, 2008). 3. von Balmoos, R., Harris, D. H. & Magee, J. S. in Handbook of Heterogeneous Catalysis (eds Ertl, G., Kno¨zinger, H. & Weitkamp, J.) 1955–1983 (Wiley-VCH, 1997). 4. Vermeiren, W. & Gilson, J-P. Impact of zeolites on the petroleum and petrochemical industry. Top. Catal. 52, 1131–1161 (2009). 5. Newsam, J. M. The zeolite cage structure. Science 231, 1093–1099 (1986). 6. Rigutto, M. in Zeolites and Catalysis: Synthesis, Reactions and Applications ˇ ejka, J., Corma, A. & Zones, S. I.) 547–584 (Wiley-VCH, 2010). (eds C 7. Corma, A. Inorganic solid acids and their use in acid-catalyzed hydrocarbon reactions. Chem. Rev. 95, 559–614 (1995). 8. Bomgardner, M. M. Using pore power. Chem. Eng. News 89, 20–21 (2011). 9. Rawlence, D. J. & Gosling, K. FCC catalyst performance evaluation. Appl. Catal. 43, 213–237 (1988). 10. Chao, K. J., Lin, L. H., Ling, Y. C., Hwang, J. F. & Hou, L. Y. Vanadium passivation of cracking catalysts by imaging secondary ion mass spectrometry. Appl. Catal. A 121, 217–229 (1995). 11. Kugler, E. L. & Leta, D. P. Nickel and vanadium on equilibrium cracking catalysts by imaging secondary ion mass spectrometry. J. Catal. 109, 387–395 (1988). 12. Cao, H. & Suib, S. L. Spectroscopic studies of the migration of vanadium in the model fluid catalytic cracking process. Appl. Spectrosc. 49, 1454–1462 (1995). 13. Lappas, A. A., Nalbandian, L., Iatridis, D. K., Voutetakis, S. S. & Vasalos, I. A. Effect of metals poisoning on FCC products yields: studies in an FCC short contact time pilot plant unit. Catal. Today 65, 233–240 (2001). 14. Haas, A., Suarez, W. & Young, G. W. Evaluation of metals contaminated FCC catalysts. AIChE Symposium Series 133–142 (1992). 15. Occelli, M. L., Gould, S. A. C. & Stucky, G. D. The study of the surface topography of microporous materials using atomic force microscopy. Stud. Surf. Sci. Catal. 485–492 (1994). 16. Psarras, A. C., Iliopoulou, E. F., Nalbandian, L., Lappas, A. A. & Pouwels, C. Study of the accessibility effect on the irreversible deactivation of FCC catalysts from contaminant feed metals. Catal. Today 127, 44–53 (2007). 17. Zhang, J., Campbell, R. E., Ting, A. Y. & Tsien, R. Y. Creating new fluorescent probes for cell biology. Nature Rev. Mol. Cell Biol. 3, 906–918 (2002). 18. Krutzik, P. O. & Nolan, G. P. Fluorescent cell barcoding in flow cytometry allows high-throughput drug screening and signaling profiling. Nature Methods 3, 361–368 (2006). 19. Martin, B. R., Giepmans, B. N. G., Adams, S. R. & Tsien, R. Y. Mammalian cell-based optimization of the biarsenical-binding tetracysteine motif for improved fluorescence and affinity. Nature Biotechnol. 23, 1308–1314 (2005). 20. Rajagopal, J., Anderson, W. J., Kume, S., Martinez, O. I. & Melton, D. A. Development: insulin staining of ES cell progeny from insulin uptake. Science 299, 363 (2003). 21. Darzynkiewicz, Z., Bedner, E., Li, X., Gorczyca, W. & Melamed, M. R. Laser-scanning cytometry: a new instrumentation with many applications. Exp. Cell Res. 249, 1–12 (1999). 22. Marks, K. M. & Nolan, G. P. Chemical labeling strategies for cell biology. Nature Methods 3, 591–596 (2006). 23. Karwacki, L. et al. Morphology-dependent zeolite intergrowth structures leading to distinct internal and outer-surface molecular diffusion barriers. Nature Mater. 8, 959–965 (2009). 24. Roeffaers, M. B. J. et al. Spatially resolved observation of crystal-face-dependent catalysis by single turnover counting. Nature 439, 572–575 (2006). 25. Xu, W., Kong, J. S., Yeh, Y. T. E. & Chen, P. Single-molecule nanocatalysis reveals heterogeneous reaction pathways and catalytic dynamics. Nature Mater. 7, 992–996 (2008). 26. Naito, K., Tachikawa, T., Fujitsuka, M. & Majima, T. Single-molecule observation of photocatalytic reaction in TiO2 nanotube: importance of

30.

31. 32. 33. 34.

35.

36. 37. 38.

39.

40.

41. 42.

43. 44.

molecular transport through porous structures. J. Am. Chem. Soc. 131, 934–936 (2009). De Cremer, G., Sels, B. F., De Vos, D. E., Hofkens, J. & Roeffaers, M. B. J. Fluorescence micro(spectro)scopy as a tool to study catalytic materials in action. Chem. Soc. Rev. 39, 4703–4717 (2010). Chen, P. et al. Single-molecule fluorescence imaging of nanocatalytic processes. Chem. Soc. Rev. 39, 4560–4570 (2010). Tachikawa, T. & Majima, T. Single-molecule, single-particle fluorescence imaging of TiO2-based photocatalytic reactions. Chem. Soc. Rev. 39, 4802–4819 (2010). Weckhuysen, B. M. Chemical imaging of spatial heterogeneities in catalytic solids at different length and time scales. Angew. Chem. Int. Ed. 48, 4910–4943 (2009). Denk, W., Strickler, J. H. & Webb, W. W. Two-photon laser scanning fluorescence microscopy. Science 248, 73–76 (1990). Roeffaers, M. B. J. et al. Morphology of large ZSM-5 crystals unraveled by fluorescence microscopy. J. Am. Chem. Soc. 130, 5763–5772 (2008). Seebacher, C. et al. Visualization of mesostructures and organic guest inclusion in molecular sieves with confocal microscopy. Adv. Mater. 13, 1374–1377 (2001). Kox, M. H. F., Mijovilovich, A., Sa¨ttler, J. J. H. B., Stavitski, E. & Weckhuysen, B. M. The catalytic conversion of thiophenes over large H-ZSM-5 crystals: an X-ray, UV/vis, and fluorescence microspectroscopic study. ChemCatChem 2, 564–571 (2010). Kox, M. H. F. et al. Label-free chemical imaging of catalytic solids by coherent anti-Stokes Raman scattering and synchrotron-based infrared microscopy. Angew. Chem. Int. Ed. 48, 8990–8994 (2009). Rautiainen, E., Pimenta, R., Ludvig, M. & Pouwels, C. Deactivation of ZSM-5 additives in laboratory for realistic testing. Catal. Today 140, 179–186 (2009). Bendiksen, M., Tangstad, E. & Myrstad, T. A comparison of laboratory deactivation methods for FCC catalysts. Appl. Catal. A 129, 21–31 (1995). Psarras, A. C., Iliopoulou, E. F., Kostaras, K., Lappas, A. A. & Pouwels, C. Investigation of advanced laboratory deactivation techniques of FCC catalysts via FTIR acidity studies. Microporous Mesoporous Mater. 120, 141–146 (2009). Mitchell, B. R. Metal contamination of cracking catalysts. 1. Synthetic metals deposition on fresh catalysts. Ind. Eng. Chem. Prod. Res. Dev. 19, 209–213 (1980). Rust, M. J., Bates, M. & Zhuang, X. Sub-diffraction-limit imaging by stochastic optical reconstruction microscopy (STORM). Nature Methods 3, 793–795 (2006). Betzig, E. & Trautman, J. K. Near-field optics: microscopy, spectroscopy, and surface modification beyond the diffraction limit. Science 257, 189–195 (1992). Chen, N. Y., Mitchell, T. O., Olson, D. H. & Pelrine, B. P. Irreversible deactivation of zeolite fluid cracking catalyst. 2. Hydrothermal stability of catalysts containing NH4Y and rare earth Y. Ind. Eng. Chem. Prod. Res. Dev. 16, 247–252 (1977). Chen, D. et al. Acidity studies of fluid catalytic cracking catalysts by microcalorimetry and infrared spectroscopy. J. Catal. 136, 392–402 (1992). Farneth, W. E. & Gorte, R. J. Methods for characterizing zeolite acidity. Chem. Rev. 95, 615–635 (1995).

Acknowledgements The authors thank Albemarle Catalysts for financial support and for providing catalyst materials, catalytic performance data and part of the bulk characterization data. J.R.M. acknowledges the ACTS-Aspect program for funding. The authors thank F. Soulimani (Utrecht University) for help with the IR measurements, M. Versluijs-Helder (Utrecht University) for the XRD measurements, U. Deka (Utrecht University) for the calculations of the unit cell sizes, A. Ruppert (Technical University of Lodz) for the design of Fig. 1a and the graphical abstract and J. Francis (Albemarle Corporation) for fruitful discussions.

Author contributions I.L.C.B. and J.R.M. contributed equally to this work. They carried out the experiments, the statistical analysis and wrote the manuscript. W.V.K. contributed to the catalytic activity tests and discussion thereof and participated in manuscript preparation. D.B., J.A.B. and E.T.C.V. contributed to the discussion of the results. B.M.W. designed and directed the research, and contributed to the preparation and writing of the manuscript.

Additional information The authors declare no competing financial interests. Supplementary information accompanies this paper at www.nature.com/naturechemistry. Reprints and permission information is available online at http://www.nature.com/reprints. Correspondence and requests for materials should be addressed to B.M.W.

NATURE CHEMISTRY | VOL 3 | NOVEMBER 2011 | www.nature.com/naturechemistry

© 2011 Macmillan Publishers Limited. All rights reserved.

867